J Cancer Res Clin Oncol (1992) 118:502-508

Cancer Nesearch Clinical 9 9 Springer-Verlag 1992

Characterization of insulin-like growth factor I receptors and growth effects in human lung cancer cell lines Martin Rotsch, Michael Maasberg, Cebrail Erbil, Gabriele Jaques, Ursula Worsch, and Klaus Havemann Philipps-UniversityMarburg, Department of Internal Medicine,Divisionof Haematology/Oncology/Immunology, Baldingerstrasse, W-3550 Marburg, Federal Republic of Germany Received 6 March 1991/Accepted21 February 1992

Summary. Small-cell lung cancer (SCLC) and non-smallcell lung cancer (NSCLC) cell lines were studied for insulin-like growth factor I (IGF-I) receptor expression and with regard to the influence of IGF-I on cell proliferation. IGF-I receptors on the cells were characterized by competitive binding assays, chemical crosslinking and northern blot hybridization of IGF-I receptor mRNA. All SCLC and NSCLC cell lines showed specific IGF-I binding sites with an affinity (KD) of 0.69 5.21 nM. The amount of binding sites ranged from 59 fmol/mg to 1230 fmol/mg protein. The IGF-I binding was inhibited by the IGF-I receptor antibody (e-IR-3). Northern blot hybridization indicated that IGF-I receptor mRNA was being produced by all SCLC and NSCLC cell lines. We used the soft-agarose clonogenic assay to evaluate the influence of IGF-I on the in vitro proliferation of the cells. Our results have shown that IGF-I stimulates the growth of all tested cell lines ranging from a factor of 1.6 to 4.2 in SCLC and from 1.1 to 2.7 in NSCLC. The data indicate that the IGF-I receptor thus appears to be the common pathway for the mitogenic activity of IGF-I and IGF-II with regard to human lung cancer cells. Key words: Small-cell lung cancer - Non-small-cell lung cancer - IGF-I receptor

Introduction The insulin-like growth factor I (IGF-I), a polypeptide hormone having structural homology with IGF-II and insulin, was purified and sequenced from human serum (Humbel and Rinderknecht 1978). IGF-I is a single-chain peptide, its primary structure is similar to that ofproinsulin, which consists of an A-chain-like part (A domain), a Abbreviations. IGF-I, insu1~n-tikegrowth factor I; SCLC,smaIl-ceI1 lung cancer; NSCLC, non-small-celllung cancer Offprint requests to: M. Rotsch

B-chain-like N-terminal part (B domain) and a connecting peptide (C domain). The A-domain and B-domain are cross-linked by three disulphide bridges. IGF-I and IGF-II consists of 70 and 67 amino acid residues respectively. The molecular masses are 7649 Da for IGF-I and 7471 Da for IGF-II (Humbel 1984). Whereas insulin plays an important role in the regulation of a variety of metabolic processes, the IGF appear to be more potent in promoting growth. IGF-I mediates the growth-promoting actions of the growth hormone on skeletal cartilage and other tissues (Van Wyk 1984). The action of IGF-II in vivo is less clear, but studies of rat IGF-II expression suggest that it has a function with regard to fetal and early neonatal development (Morgan et al. 1987). In vitro, there are cell culture systems in which IGF per se in physiological concentrations can mediate mitogenic effects, for example, in BALB/c 3T3 mouse fibroblasts (Pledger et al. 1977). The physiological actions of IGF are mediated through specific interactions with cellsurface receptors. The type I receptor has a higher affinity for IGF-I than for IGF-II and also binds insulin with low affinity (Nissley and Rechler 1984). This receptor with a molecular mass of 300-350 kDa is a membrane glycoprotein consisting of two subunits of 135 kDa and two subunits of 90 kDa that are connected by disulphide bonds to form a heterotetrameric receptor complex (Chernausek et al. 1987). The IGF-II receptor is to be distinguished from the IGF-I and the insulin receptor. It possesses a high affinity for IGF-II, a low affinity for IGF-I and does not bind insulin (Kasuga et al. 1981). The primary molecular structures of the IGF-I and IGF-II receptor have been determined from cloned cDNA (Morgan et al. 1987; Ullrich et al. 1986). The concept of autocrine and paracrine growth stimulations has become an attractive model in cancer research (Sporn and Todaro 1980). The malignant transformation of cells may be accompanied by the responsiveness of autocrinely or paracrinely produced growth factors and may be linked with the presence and surface density of receptors for these growth factors. Recently it has been reported that human breast cancer cells (Furlanetto and Di

503

Carlo 1984), human leukaemia cells (Vetter et al. 1986) and human choriocarcinoma cells (Ritvos et al. 1988) express receptors for IGF-I and that IGF-I has a mitogenic effect on these cells. In the present study we will show that in permanent cell lines of small-cell (SCLC) and nonsmall-cell lung cancer (NSCLC) IGF-I receptors can be found. Characteristics of the IGF-I receptor have been established by competitive ligand binding, chemical cross-linking techniques and Northern blot analysis of the IGF-I receptor mRNA. An additional goal of this study has been to determine the clonogenic effect of IGFI on these cell lines. Materials and methods Cell lines. The cell lines used in this study were established at the Philipps University Medical Center, Marburg, F R G (Bepler et al. 1987, 1988), at the National Cancer Institute, Bethesda, USA (Carney et al. 1985), at the Dartmouth-Hitchcock Medical Center, Hannover, USA (Pettengill et al. 1980) and at the Uppsala Medical Center, Sweden (Bergh et al. 1981, 1985). For details see Table 1. MCF-7, HL60, Daudi and A431 cell lines were obtained from the European Collection of Animal cell Cultures (ECACC), Salisbury, UK. SCLC cell lines grow as floating cell aggregates, whereas all other tung cancer cell lines grow substrate-adherently. They were kept in RPMI-1640 medium (no.041-1876; Gibco, Paisly, UK), which was supplemented by 10% fetal bovine serum (no. 011-6290; Gibco) in a well-humidified atmosphere of 5% CO2 and 37 ~ C and showed no contamination with Mycoplasma. The cells were collected during logarithmic growth phase 48 h following medium change before the tests. 125I-IGF-I binding assay. NSCLC: cells in monolayer cultures were collected with a cell scraper, washed tiwce with phosphate-buffered saline (PBS) and resuspended in RPMI medium with 0.1% (w/v) bovine serum albumin. SCLC: cells were washed twice with PBS and resuspended in RPMI medium with 0.1% bovine serum albumin. All cells were adjusted to (4-6) x 10 6 cells/ml. The IGF-I binding sites were determined by means of a centrifugation binding assay. Labelled human recombinant 125I-Tyr-IGF-I (Amersham International, Buckinghamshire, UK; no. IM-172, specific activity 2000 Ci/ mmol) and unlabelled IGF-I (Amersham, no. A R N 4010) were added to the cell suspension. For other competitive experiments IGF-II (Lilly-Inc., FRG), porcine insulin (Sigma, no. 1-3505) and

Table 1. Origin and designation of cell lines

Marburg

NCI

Dartmouth

Uppsala

Small-cell lung cancer SCLC-16HC SCLC-16HV SCLC-21H SCLC-22H SCLC-24H SCLC-86MI Non-small-cell lung EPLC-32MI EPLC-65H LCLC-97TM1 LCLC-103H MSTO-211H

NCI-H60 NCI-H69 NCI-H82 NCI-HI46 NCI-H526 NCI-N592 NCI-H841 cancer NCI-H23 NCI-HI25 A549

DMS-79

U-1752 U-1810

the IGF-I receptor antibody (c~-IR-3) a gift of Dr. S. Jacobs, Wellcome Research Lab. N.C., USA) were used. After a 90-min incubation at 4 ~ C, unbound radioactivity was removed by centrifugation at 10000 g for 2 min. The cells were washed twice in RPMI-1640 medium with 0.1% bovine serum albumin. The bound radioactivity of the cell pellets was counted in a gamma counter (LKB, Sweden). The maximal binding (Bmax) per milligram protein and the dissociation constant (Kd) were calculated with the help of Scatchard plot analysis (Bennett 1978). The protein content of cell pellets was measured with a Bio-Rad standard protein assay (Coomassie brillant blue G-250; Bio-Rad laboratories, FRG).

RNA extraction and Northern hybridization. RNA was prepared by the method described by Chirgwin et al. 1979. The polyadenylated m R N A fraction was isolated by oligo(dT)-cellulose chromatography. A 15-p,g sample of RNA was electrophoresed into a 1% glyoxal/agarose gel and blotted onto nitrocellulose. The nitrocellulose filters were hybridized with a 3/P-labelled cDNA probe in 50% formamide, 5 x standard saline citrate, 25 m M sodium phosphate (pH 6.5), 100 gg/ml denatured salmon DNA, 2 x Denhardt's solution at 42 ~ C for 15-20 h. After the incubation, the filters were extensively washed in 0.2 x standard saline citrate, 0.1% sodium dodecyl sulphate at 50 ~ C and then autoradiographed. The human IGF-I receptor cDNA was a gift from A. Ullrieh Genentech Inc., San Francisco. A 730-base-pair EcoRI fragment of the IGF-I receptor cDNA insert ofplasmid pIGF-I-R.8 was used to probe for IGFI receptor mRNA. Chemical cross-linking of 125I-labelled IGF-I to cultured cells. The cross-linking of 125I-IGF-I was performed on membrane preparations of the cells. Cells were homogenized with a Dounce-type glass homogenizer in 5 m M TRIS, 1 m M EDTA, 0.25 M sucrose (pH 8.0) with 1000 U aprotinin/ml. The homogenate was centrifuged at 1000 g for 5 min. The supernatant was centrifuged at 33000 g for 10 min in a Ti70 rotor at 4 ~ C. The second step was repeated twice. The pellet was resuspended in 25 m M HEPES, 10 n M CaC12, 0.1% Triton X-100 and 0.1% bovine serum albumin (pH 7.4). A 50-gg sample of membrane protein with in a final volume of 500 gt was incubated with 50 pg iodinated IGF-I for 90 min at 4 ~ C, then 50 ng unlabelled IGF-I, 50 ng IGF-II, 500 ng insulin and 500 ng IGF-I receptor antibody (~ IR-3) were added for competition with the lzsI-IGF-I binding. Bound peptide was separated from free peptide by adding 10 gl 1.25% bovine immunoglobulin and 400 gl polyethylene glycol 16% w/w. The samples were centrifuged for 20 rain, 3000 g, 4 ~ C. The pellets were resuspended in 50 gl buffer (0.1 MHEPES, 0.15 MNaC1, 5 mMKC1, 8 mMglucose and 1% Triton X-100, pH 8.0). The cross-linking was implemented by adding 2 gl dissuccinimidyl suberate (Pierce Clinical Co., Rochford, USA) to the membrane suspension (final concentration: 0.5 m M disucciuimidyl suberate). The samples were incubated at room temperature for 15 min and the reaction was quenched by adding 25 ~tl 1 M TRIS, 0.2 M EDTA, pH 6.8 and 75 gl gel sample buffer containing 5% sodium dodecyl sulphate (SDS). Samples were boiled and subjected to 3%-15% polyacrylamide gel electrophoresis, they were then dried and examined autoradiographically. Soft-agarose clonogenic assay. The soft-agarose clonogenic assay was used to determine the colony-forming efficiency of the SCLC cell line in the presence of different concentrations of IGF-I (Carney et al. 1984). A base layer of I ml 0.5% agarose (Seakem LE, Rockland, Me., USA in RPMI-1640) medium supplemented with 100 n M transferrin (Sigma no. T-2252), 30 n M sodium selenite (Sigma no. T-9133) and 0.25% bovine serum albumin was prepared in 35-mm dishes and allowed the time to harden at room temperature. Cells were collected during logarithmic growth phase, washed twice with RPMI-1640 medium, and a single-cell suspension was prepared mechanically. The number of viable cells was determined with the help of trypan blue uptake (0.1% w/v solution). A preparation was made of a top-layer solution consisting of the respective growth medium, 0.3% w/v agarose, IGF-I in different concentra-

504 tions ranging from 1 n M to 500 nM, and viable cells at a concentration of 104 cells/ml. Dishes were incubated at 37 ~ C in a humidified atmosphere of 8% COz/92% air. The colonies that grew were counted using an Omnicon image analysis system (Bausch and Lomb, Rochester, N.Y., USA). The growth rate was evaluated by the quotient of colonies growing in the presence of IGF-I (N) versus those growing in the absence of IGF-I (No). The growth rate No in the absence of IGF-I was set as 1.

BIBo

"

,

"

l\.

k

0o,1

\

8....

207,~o~/mg

Ko=,s0nM

000i \ . 0'02i

Results

0,5-

Binding of 125I-IGF-I on cultured cells The binding of 125I_IGF_I to cultured cells was studied as a function of time at temperatures of 4, 20, and 37 ~ C. The rate at which steady-state binding was achieved varied with the incubation temperature. At 4 ~ C the maximum binding occurred after an incubation period of 90 min. At 37 ~ C maximum binding was achieved after a 15-min incubation. After longer incubations the cellbound radioactivity decreased down to 65 % of the maximum value at 37 ~ C (data not shown). Taking these results as a basis, the ~25I-IGF-I binding assay was performed with an incubation period of 90 min at 4 ~ C. Different SCLC and N S C L C cell lines were studied in displacement experiments with a constant amount of labelled IGF-I (50 pM) and various quantities of unlabelled IGF-I, IGF-II and porcine insulin in the concentration range 0.1-100 nM. The effect of increased concentrations of unlabelled IGF-I, IGF-II and insulin on tzsI-IGF-I binding is shown in Fig. 1 for the NCI-H69 cell line. Judg-

0,0

5() fmollOOmgProtein 2~00 ~

, ,

0;5

I

I0

1CIOnM

IGF-I "

Fig. 2. Competition of labelled and unlabelled IGF-I for binding to SCLC cell line NCI-H69. Indicated amounts of unlabelled IGF-I and 12sI-IGF-I were added simultaneously to the cells. Ordinate: bound IGF-I (B), bound IGF-I in absence of unlabelled IGF-I (B0); abscissa: unlabelled IGF-I (nM). Inserts: Scatchard plot of binding data

a/ao

IgO

1

BIB0 1,0~

O

~ 4

i :

0,5

o Insulin

lb

1;OnM

1'o

,;~

~;~

~;'~

Fig.3. Inhibition of lzsI-IGF-I binding to small-cell lung cancer (SCLC) cell lines NCI-H82 and SCLC-24H by the monoclonal IGFI receptor antibody eIR-3. The indicated concentrations of elR-3 (ng/ml) and a mouse IgG preparation as control were preincubated for 30 min at 20 ~ C and then incubated with 30000 cpm t25I-IGF-I for 90 min at 4 ~ C. The unbound radioactivity was removed by centrifugation

IGF-I

z~ I G F - II

~176 ' ols

;

--

Fig.l, Displacement of ~25I-labelled insulin-like growth factor I (~25I-IGF-I) binding to the NCI-H69 cell line. NCI-H69 cells were incubated with 50 pg x25I-IGF-I and the indicated concentrations of unlabelled peptides for 90 rnin at 4 ~ C. Each point represents the mean of triplicate measurements

ing from half-maximal inhibitory concentrations, IGF-II has a comparable potency to IGF-I whereas insulin was found to be less potent. A 50% displacement was achieved with unlabelled insulin at a concentration of approximately 100 nM, i.e. a concentration that is ten times higher than that of IGF-I or IGF-II. Figure 2 illustrates the Scatchard analysis of 12sIIGF-I binding of the NCI-H69 cell line. The binding data for the NCI-H69 cell line are consistent with a single,

505 high-affinity binding site with Kd = 1.80 n M and a r e a x = 207 fmol/mg protein. Evaluations of all cell lines tested are summarized in Table 2. IGF-I binding sites were found in all tested SCLC and N S C L C cell lines including MCF-7 and HL60 used as controls. The Bmax ranged Table 2, Insulin-like growth factor I binding sites in human cancer cell lines Cell line SmaU-cell carcinoma Classic SCLC-16HC SCLC-22H SCLC-24H SCLC-86M1 NCI-H69 NCI-H146 NCI-N592 Variant SCLC 16HV SCLC-21H NCI-H82 NCI-N526 DMS-79 Squamous cell carcinoma EPLC-32M1 EPLC-65H U-1752 Adenocarcinoma NCI-H23 NCI-H125 A549 Large-cell carcinoma LCLC-97MI LCLC--103H U-1810 Mesothelioma MSTO-211H Breast carcinoma MCF-7 Myeloid leukemia HL-60

Maximal binding (fmol/mg)

Kd (nM)

from 59 fmol/mg to 1230 fmol/mg protein and the Ko from 0.69 n M t o 5.21 nM. A marked difference in the dissociation constant and in the maximal amounts of IGF-I bound has not been found for N S C L C and SCLC cell lines. The effects of eIR-3 on the binding of ~2sI-IGF-I were tested on SCLC-24H and NCI-H82 cell lines. The antibody inhibited the binding of labelled IGF-I to either cell line (Fig. 3). The antibody eIR-3 binds to and inhibits IGF-I binding to intact cells, the epitope with which it reacts is the e subunit of the I G F - I receptor.

IGF-I receptor: Northern analysis 309 330 317 186 207 164 272

1.25 1.02 0.89 1.33 1.80 0.61 3.05

532 1 121 1 230 400 131

5.21 3.19 3.08 2.88 2.80

721 280 361

2.43 0.75 4.64

247 110 304

0.74 0.94 1.67

59 345 261

I. 10 1.53 0.99

127

1.48

483

2.72

213

2.24

In order to verify that IGF-I receptors were synthesized by SCLC and N S C L C cell lines, we extracted the total cellular R N A and examined the expression of the IGF-I receptor gene through Northern blot hybridization. The results indicate that IGF-I receptor m R N A was found in all SCLC and N S C L C cell lines with the exception of U1810, which yielded fainter signals. Hybridization with the c D N A probe yielded signals of 11 x 103 bases (kb) and 7 kb in accordance with the data of others (Ullrich et al. 1986). As negative control the Daudi cell line was used. An autoradiograph of a Northern blot is shown in Fig. 4. The level of expression varied considerably between different cell lines. There was no correlation between the amount of binding sites (Bmax) and m R N A expression.

Chemical cross-linking of a25I_IGF_I to cultured cells In order to establish further characteristics o f the IGF-I receptors on the celt lines an affinity labelling technique, similar to that used for the characterisation o f the IGF-I receptor in other tissues, was employed (Kasuga et al. 1981; Furlanetto and Di Carlo 1984; Ritvos et al. 1988). Figure 5 shows the autoradiogram of receptor complexes of NCI-H82 cross-linked to 12sI-IGF-I analysed on SDS/ polyacrylamide gel electrophoresis under non-reducing conditions. In the cell line NCI-H82, IGF-I was specifi-

Fig. 4. Northern blot analysis of IGF-I receptor mRNA of SCLC and non-small cell lung cancer (NSCLC) cell lines. A 730-bp EcoRI fragment of pIGF-I-R.8 was used. An IGF-I-receptor-specific transcript of 7 kb and 11 kb was detected. The mobilities of 28-8 (5.0 kb) and 18-S (2.1 kb) ribosomal RNAs are indicated on the right. A, SCLC-16HV; B, SCLC-22H; C, SCLC-24H; D, NCI-H69; E, NCIH82; F, SCLC-16HC; G, SCLC-86MI; H, DMS79; I, NCI-H23; J, EPLC-65H; K, EPLC-32MI; L, LCLC-103H; M, NCIH125; N, LCLC-79TM1; O, A549; P, U-1810; Q, HL-60; R, A431; S, Daudi

506 body (elR-3) (lane 6). Normal mouse immunoglobulin was found to have no effect (lane 7). Insulin, at high concentrations (500 ng), also binds to the IGF-I receptor. In the presence of porcine insulin the binding of labelled IGF-I to the specific band decreased (lane 5).

Growth-stimulating effects of lGF-I on cultured cells IGF-I stimulated the growth of 10/l ] SCLC cell lines in the soft-agarose clonogenic assay by a factor N/No of 1.6-4.2 in serum-free medium. IGF-I had only a moderate growth effect on NCI-H526 (N/No = 1.4) (Fig. 6). The growth stimulation was found within a range of 5500 riM; 5/10 of the cell lines were stimulated by a low factor (N/No = 1.6-2.0; data not shown). The stimulatory effect of IGF-I on NSCLC cell lines was tested on EPLC65H, LCLC-97M1 and LCLC-103H. The maximum stimulation was found in LCLC-97M1, with N/No =2.7 at IGF-I concentrations from 5 nM to 500 nM (data not shown). Fig. 5. Autoradiogram showing the size and specificity of 125I.IGF_I cross-linked complexes on the small-cell lung cancer cell line NCIH82 under non-reducing conditions. Cells were incubated with 125IIGF-I in absence (lane 2) and presence of unlabelled IGF-II (lane 3), IGF-[ (lane 4), insulin (lane 5), ~IR-3 (lane 6) and non-specific immunoglobulin (lane 7). Samples were cross-linked with 0.5 m M disuccinimidyl suberate, solubilized in gel sample buffer, and gels were autoradiographed as described in Materials and methods. The numbers on the left show the position of the molecular mass marker proteins (lane 1). A 5%-25% polyacrylamide gradient was used

5 ~

4~ o

Z

Z

3-

21-

I

i

0

5

9 'NOI-H82 [ ] NCI-H526

i

t

10 50 IGFI (nM) [ ] SCLO-24H [ ] SCLC-22H

i

i

100

500

[ ] SOLC-86M1

Fig. 6. Effect of IGF-I in different concentrations on the proliferation of five SCLC cell lines. The relative cell proliferation N/N o represents the number of colonies in triplicate plates of IGF-I-treated (N) versus untreated (No) cells

caUy bound to a protein of high molecular mass in the 300-350 kDa range. With regard to demonstrating that the prominent band represented specific IGF-I binding, the specificity of reaction was proven by competition with an excess of unlabelled IGF-I (lane 4). Unlabelled IGF-I erased the radioactivity from the band. In the presence of IGF-II the t25I-IGF-I binding to the receptor protein band was similarly decreased (lane 3). For additional characterisation of the IGF-I binding, the ~25IIGF-I binding was displaced by an IGF-I receptor anti-

Discussion

IGF-I and insulin show considerable sequence homology. Each peptide binds to its own receptor with high affinity and to the other receptor with less affinity. For small-cell lung cancer cells in vitro, insulin has been found to be one of the potent growth stimulators although the optimal proliferation was found in supraphysiological concentrations (Simms et al. 1980). It has been suggested that the mitogenic effect of insulin is not mediated by the insulin receptor (King et al. 1980). This paper demonstrates that IGF-I stimulates cell growth in 10/11 tested SCLC cell lines and in 2 NSCLC cell lines. In all cell lines examined, IGF-I stimulated the cell growth in the absence of other serum components. In our study the receptor binding assays demonstrate highaffinity binding sites for IGF-I in all lung cancer cell lines examined, c~IR-3, a monoclonal antibody, binds with an epitope on the external site of the IGF-I receptor; a weak cross-reaction with insulin receptors was described (Jacobs et al. 1986). The antibody particularly inhibits the IGF-I binding to its receptor. According to our results the cdR-3 antibody inhibits IGF-I binding in the lung cancer cell lines. An example is illustrated in Fig. 3. An evaluation of our results with regard to the binding studies and the growth effects leads us to assume a highaffinity IGF-I receptor in the lung cancer cell lines. Our data, on the whole, confirm the results published by Nakanishi et al. (1988), who reported similar effects of IGF-I and cdR-3 on the lung cancer cell lines NCI-H82, NCI-H209, NCI-H345 and NCI-N417. For a classification of the IGF-I binding sites in lung cancer cells as IGF-I receptor, the cell lines were tested through Northern blot hybridization for IGF-I receptor mRNA expression. We have used a cDNA probe with high specificity for the c~ subunit of the IGF-I receptor. The Northern blot hybridization with this IGF-I receptor probe showed an expression of IGF-I receptor m R N A in 17/19 of the cell lines examined. Analysis of IGF-I recep-

507 tor m R N A size and expression revealed 7-kb and 11-kb m R N A s in h u m a n tissues. The two I G F - I receptor m R N A s were identified in h u m a n term placenta (Ullrich et al. 1986) and detected in the h u m a n leukemic cell line HL60 (Kellerer et al. 1990). Nakanishi et al. (1988) analysed the cell lines N C I N417 and N C I - H 3 4 5 for the expression of I G F - I precursor molecules through Western blot analysis. Their study revealed a 16-kDa band in both cell types but no natural I G F - I was detected in the cell lines tested. In evaluating these results and in finding that I G F - I stimulates the growth of lung cancer cell lines, we conclude that I G F - I can function as an autocrine or paracrine growth factor for lung cancer cells, which fulfills all criteria of the autocrine/paracrine model of t u m o u r growth. U p till now only gastrin-releasing peptide (a m a m m a l i a n equivalent of bombesin) has been suggested to be an autocrine growth factor for SCLC. I G F - I and I G F - I I are chemically similar peptides and both of them stimulate growth by binding to the type I I G F receptor (E1-Badry et al. 1991). The expression of a high level of I G F - I I in a n u m b e r of h u m a n embryonal and adult tumours has raised the possibility that I G F - I I might contribute to the neoplastic proliferation of tum o u r cells. I G F - I I secretion by lung cancer cells has not been demonstrated yet. Another important finding for the autocrine/paracrine growth stimulation is the production of I G F - b i n d ing proteins by most of the SCLC cell lines (Jaques et al. 1989). Jaques et al. were able to demonstrate the release of I G F - b i n d i n g proteins by 6/9 of the SCLC cell lines in serum-free media. The binding proteins specifically bind I G F - I and I G F - I I and differ from the known binding protein: IBP-1. In conditioned media these SCLC cell lines showed a specific binding for I G F - I in the molecular mass range of 24-32 kDa. In our experiments a high-molecular-mass band was preferentially labelled when 125I_ I G F - I was cross-linked with SCLC cells and analysed by gel electrophoresis. Our results have demonstrated in cross-linking experiments that the high-molceular-mass band represents on I G F - I receptor. However, a low-molecular-mass band, corresponding to the binding proteins, was not found under the assay conditions used. The reason for the lack of the low-molecular-mass band is to be found in the low concentration of binding proteins in the m e m b r a n e preparation of t u m o u r cells used. The function of these binding proteins needs further clarification. They could be important mediators in the paracrine pathway by supporting the tumour cells with I G F produced in the surrounding tissue.

Acknowledgements. We thank E. Wiesing and A. Immel for technical assistance. The work was supported by the Sonderforschungsbereich 215 of the German Research Society.

References Bepler G, Jaques G, Neumann K, Aumfiller G, Gropp C, Havemann K (1987) Establishment, growth properties and morphological characteristics of permanent human small cell lung cancer cell lines. J Cancer Res Clin Oncol 113:31 Bepler G, Kiefer P, Havemann K, Beisenherz K, Jaques G, Gropp C, Haeder M (1988) Characterization of the state of differentiation of six newly established human non-small cell lung cancer cell lines. Differentiation 37:158 Bennett JP Jr (1978) Methods in binding studies. In: Yamamura HI (ed) Neurotransmitter receptor binding. Raven P, New York, pp 57-89 Bergh J, Nilson K, Zech L, Giovanella B (1981) Establishment and characterisation of a continuous lung squamous cell carcinoma cell line (U- 1752). Anticancer Res 1:317 Bergh J, Ekam R, Giovanella B, Nielson K (1985) Establishment and characterisation of cell lines from human small cell and large cell carcinoma of the lung. Acta Pathol Microbiol Immunol Scand [A] 93:133 Carney DN, Gazdar AF, Bunn PA, Guccion JG (1984) Demonstration of stem cell nature of clonogenic tumor cell from lung cancer patients. Stem Cell 1:149 Carney DN, Gazdar AF, Bepler G, Guccion JG, Marangos PJ, Moody TW, Zweig MH, Minna JD (1985) Establishment and identification of small cell lung cancer cell lines having classic and variant features. Cancer Res 45:2913 Chernausek SD, Beach DC, Banach W, Sperling MA (1987) Characteristics of hepatic receptors for somatomedin-C/insulin-like growth factor I and insulin in the developing human. J Clin Endocrinol Metab 64:737 Chirgwin JM, Przybyla AE, Mac Donald RJ (1979) Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18:5299 E1-Badry OM, Helman LJ, Chatten J, Steinberg SM, Evans AE, Israel MA (1991) Insulin-like growth factor II-mediated proliferation of human neuroblastoma. J Clin Invest 87:648 Furlanetto RW, Di Carlo JN (1984) Somatomedin-C receptors and growth effects in human breast cells mainteined in long-term culture. Cancer Res 44:2122 Humbel RE (1984) Insulin-like growth factors, somatomedins and multiplication stimulation activity: chemistry. In: Li CH (ed) Hormonal proteins and peptides, vol 12. Academic Press, Orlando, pp 57-79 Humbel RE, Rinderknecht E (1978) The amino acid sequence of human insulin like growth factor I and its structural homology with proinsulin. J Biol Chem 258:2769 Jacobs S, Cook S, Svoboda ME, Van Wyk J (1986) Interaction of the monoclonal antibodies aiR-1 and aiR-3 with insulin and somatomedin-C receptors. Endocrinology 118:223 Jaques G, Kiefer P, Rotsch M, Hennig C, G6ke R, Richter G, Havemann K (1989) Production of insulin-like growth factor binding proteins by small cell lung cancer cell lines. Exp Cell Res 184:396 Kasuga M, Van Obberghen E, Nissely SP, Rechler MM (1981) Demonstration of two subtypes of insulin-like growth factor receptors by affinity crosslinking. J Biol Chem 256:5305 Kellerer M, Obermaier K, Ermel B, Wallner U, Haring HU (1990) An altered IGF-I receptor is present in human leukemic cells. J Biol Chem 265:9340 King GL, Kahn CR, Rechler MM, Nissley SP (1980) Direct demonstration of separate receptors for growth and metabolic activities of insulin and multiplication-stimulating activity (an insulin-like growth factor) using antibodies to the insulin receptor. J Clin Invest 66:130 Morgan DD, Eckmann JC, Standring DN, Fried VA, Smith MC, Roth RA, Rutter JW (1987) Insulin-like growth factor II receptor as a multifunctional binding protein. Nature 329:301 Nakanishi Y, Mulshine JL, Kasprzyk PG, Maneckjee R, Avis I, Treston AM, Gazdar AF, Minna JD, Cuttitta F (1988) Insulin-

508 like growth factor I can mediate autocrine proliferation of human small cell lung cancer cell lines in vitro. J Clin Invest 82:354 Nissley SP, Rechler MM (1984) Insulin-like growth factors: biosynthesis, receptors and carrier proteins. In: Li CH (ed) Hormonal proteins and peptides, vol 12. Academic Press, Orlando, pp 127203 Pettengill OS, Sorenson GD, Wurster-Hill DM, Curphey T, Noll WW, Cate CC, Maurer LH (1980) Isolation and growth characteristics of continuous cell lines from small cell carcinoma of the lung. Cancer 45:906 Pledger W J, Stiles LD, Autoniades N, Scher CD (1977) Induction of DNA synthesis in BALB/c3T3 cells by serum components: reevaluation of the commitment process. Proc Natl Acad Sci USA 74:4481 Ritvos O, Rutanen EM, Pekonen F, Jalkonen J, Suikkari AM, Ranta T (1988) Characterization of functional type I insulin-like growth factor receptors from human choriocarcinoma cells. Endocrinology 122:395

Simms E, Gazdar AF, Arams PG, Minna JD (1980) Growth of human small cell (oat cell) carcinoma of the lung in serum-free growth factor supplemented medium. Cancer Res 40:4356 Sporn MB, Todaro GJ (1980) Autocrine secretion and malignant transformation of cells. N Engl J Med 303:878 Ullrich A, Gray A, Tam AW, Yang-Feng T, Tsubokawa M, Collins C, Menzel W, Le Bon T, Kathuria S, Chen E, Jacobs S, Franke U, Ramachanchan J, Fujita-Yamaguchi Y (1986) Insulin-like growth factor I receptor primary structure: comparison with insulin receptor suggests structural determinants that comparison with insulin receptor suggests structural determinants that define functional specificity. EMBO J 5:2503 Van Wyk JJ (1984) The somatomedins: biological actions and physiologic contro mechanisms. In: Li CH (ed) Hormonal proteins and peptides, vol 12. Academic Press, Orlando, pp 81-125 Vetter U, Schlickenrieder JHM, Zopf J, Hartmann W, Heir W, Hitzler H, Byrne P, Gaedicke G, Heinze E, Teller WM (1986) Human leukemic cells: receptor binding and biological effects of insulin and insulin-like growth factors. Leukemia Res 10:1201